Studies of kangaroos suggest that hopping provides energy savings during locomotion at high speeds, although studies of small mammals suggest that hopping is no more economical than running. To obtain comparative data on anurans, we exercised Fowler’s toads (Bufo woodhousei fowleri, 25·8 g) on treadmills at speeds ranging from 0·09 to 0·63 km h−1 while measuring oxygen consumption , endurance or hop kinematics. The toads walked at slow speeds and hopped at fast speeds. Steady-state increased linearly with speed to a maximum of 1·17mlO2g−1h−1 at 0·27kmh−1 and was nine times the average pre-exercise . The maximum rate of oxygen consumption during treadmill exercise was comparable to previously reported for less natural exercise regimes. At speeds ⩾0·27 kmh−1, was independent of speed. At speeds ⩽0·36 kmh−1, toads moved for over 1h, but endurance decreased sharply at higher speeds. Hop rate, hop length, hop height and angle of take-off increased with speed. Hopping in B. w. fowleri was not less costly than running in other animals of similar body size and was inefficient at converting metabolic to mechanical energy. The present study suggests that hopping in toads, as in small mammals, is not economical during sustained locomotion and is most important during short bursts of high-intensity activity.

In most terrestrial vertebrates (for a review, see Bennett, 1985), the rate of oxygen consumption increases linearly with speed of locomotion to a maximum and is thereafter independent of speed. The speed at which occurs is called the maximal aerobic speed (MAS). Because is maximal and constant at speeds ⩾MAS, these speeds are increasingly supported by anaerobic metabolism and are generally not sustainable (Bennett, 1985).

Although saltatory locomotion occurs in several orders of vertebrates, the energetics of saltatory transport are understood for only a few mammalian species. Surprisingly, an early study (Dawson & Taylor, 1973) of locomotor energetics in red kangaroos, Megaleia rufa (18–28 kg), showed that the aerobic cost of salutatory transport actually decreases with increased speed. This pattern obviously contrasts sharply with the typical pattern of metabolic support for locomotion described above. Moreover, a red kangaroo hopping at rapid sustainable speeds expends less energy at a given speed than a running quadruped of similar body mass. The kangaroo’s economical locomotion at high sustainable speeds has been attributed to the storage and recovery of strain energy from elastic elements in the legs (Alexander & Vernon, 1975; Cavagna, Heglund & Taylor, 1977).

Subsequently, other workers examined the metabolic cost of saltatory locomotion in small mammals (Baudinette, Nagle & Scott, 1976; Dawson, 1976; Thompson, MacMillen, Burke & Taylor, 1980). With perhaps one exception (Dawson, 1976), the small mammal data suggest that kangaroos are exceptional and that hopping is not generally less costly than running (Thompson et al. 1980). The difference in locomotor energetics between small and large mammalian hoppers appears to be associated with morphological differences in elastic storage (Biewener, Alexander & Heglund, 1981; Emerson, 1985).

Saltatory locomotion is not restricted to mammals, but also occurs in birds, anurans and invertebrates. These other forms are useful for examining the generality of conclusions based on mammalian studies. Therefore, we examined locomotor energetics, endurance and kinematics of an anuran, Fowler’s toad (Bufo woodhousei fowleri), exercising on a treadmill at controlled speeds. We addressed the question of whether saltatory locomotion at fast sustainable speeds provides energy savings in B. w. fowleri as it does in the red kangaroo. Subsumed in this question are two ancillary issues. First, is hopping in toads less expensive than running or walking by animals of a similar body mass? Second, how efficient are toads at converting metabolic energy input into mechanical energy output? High efficiency may indicate a significant contribution from elastic storage.

Our data also bear upon amphibian activity metabolism in general. Many previous studies have attempted to measure by shocking or overturning anurans in rotating respirometer chambers (reviewed by Taigen & Pough, 1985). Only Taigen & Beuchat (1984) related the rate of chamber rotation (presumably speed of movement) to . Although these studies have revealed differences among anuran groups in activity metabolism, they have been criticized (see below) because their exercise regime did not elicit natural locomotion and may not elicit . Therefore, an additional objective of this study was to determine if locomotion on a treadmill, which resembles natural locomotion, yields a similar to that of toads in rotating respirometer jars.

Animals

Fowler’s toads, B. w. fowleri, [body mass = 25·80 ± 1·48 g (±S.D.)] were obtained at the Indiana Dunes National Lakeshore, Porter Co., IN, and from a commercial supplier. Toads were maintained at 21°C on a 14h:10h light: dark photoperiod centred on 13.00h local time. Toads were fed crickets (Acheta domestica) twice weekly until 5 days before experiments. All experimental measurements were made at 21°C.

Oxygen consumption

Measurements of during rest, sustained hopping and recovery were made in a gas-tight clear acrylic respirometer containing a variable-speed treadmill (Herreid, 1981; Full, 1986). The chamber in which the animals exercised consisted of a half-cylinder 25·4cm long, 12·7cm wide and 6·4cm tall at its highest point. Preliminary trials with unrestrained toads on a larger treadmill used for kinematic studies (see below) indicated that the toads in the respirometer had sufficient room to allow unencumbered hopping. The interior of the treadmill was moistened with 10ml of distilled water before the toads were placed inside. Toads rested within the treadmill in dim light for 45 min to1h before exercise. Resting was measured during the last 10–15 min of the rest period. The toads were then exercised for 30min at a single speed. The speeds were 0·09, 0·18, 0·27, 0·36, 0·54 km h−1. An attempt was made to exercise each toad at each speed. Toads were run only once a day and were rested for at least 2 days between trials. Trials in which toads did not locomote consistently were not used.

Oxygen consumption was measured continuously using open-flow respirometry. An air pump drew humidified air through the system at 500 ml min−1. Incurrent fractional oxygen content was compared to the excurrent fractional oxygen content with a dual-channel oxygen analyser (Ametek Applied Electrochemistry, Model S-3A). Before analysis, gas was drawn through columns of Drierite® and Ascarite® to remove water and CO2, respectively. Output from the oxygen analyser was run through an analogue-to-digital converter (Isaac 41A system, Cyborg) and into a microcomputer (Apple II+). Data were collected at 10-s intervals.

Instantaneous was calculated from and (Full & Herreid, 1983, 1984; Herreid, 1981). Given any two measurements separated by a brief interval, the flow rate and the ‘washout’ characteristics of the chamber (Feq) (the equilibrium of inside the metabolic chamber) can be calculated and substituted for in a standard equation:
formula
to estimate instantaneous Rates of oxygen consumption are reported as ml O2 at STPD.

Anaerobic metabolism is known to contribute significantly to support of activity in amphibians moving at high, non-sustainable speeds (Taigen & Beuchat, 1984; Feder, 1986; Full, 1986). Inasmuch as our primary interest was in locomotion at lower, sustainable speeds, we chose not to undertake whole-body assays of lactate production, which necessitate killing large numbers of animals.

Endurance

Endurance was measured by exercising toads to exhaustion on a motorized treadmill. Toads were placed individually on the treadmill and allowed to rest in dim light for 15–30 min. The treadmill belt was moistened and humidified air was drawn through the chamber. Treadmill speeds ranged from 0·18 to 0·63 km h−1. Several toads were prodded during the experiments to encourage movement. Prodding consisted of quickly reversing the direction of the treadmill belt and causing the toad to contact the back wall of the chamber. Prodding was required only during the first few minutes of exercise to elicit steady movement and when toads began to fatigue. At slow speeds, toads often hopped or walked indefinitely without fatigue; these trials ceased after 2 h. Otherwise, exhaustion was defined as the time when a toad (1) did not keep pace with the treadmill, (2) lost balance and turned over on its back, and (3) did not respond to three successive proddings (Full, 1986). Toads were exercised once a day and rested for at least 2 days between trials. Trials in which toads did not locomote consistently were discarded.

Kinematics: hop rate, length, height and angle of take-off

Hop rate (hopss−1), hop length (cm), hop height (cm) and angle of take-off (degrees) were determined from videotapes recorded while toads hopped on a motorized treadmill. The treadmill chamber consisted of an open-top rectangular box made of transparent acrylic 41cm long, 15 cm wide and 21·6cm high. A coordinate grid of 1-cm squares was placed on the chamber wall opposite a Panasonic WV-3230 video camera/recorder. The camera was mounted on a tripod 1 m from the treadmill. The camera was focused so that the treadmill chamber and grid filled the field of view. The toads were videotaped while moving at each of the five speeds used in the experiments. Recordings were made consecutively at these speeds and the order of speeds was randomized for each toad. Maximum height and length of hops were measured against the grid as the videotapes were advanced in 0·03–0·04 s increments. Angle of take-off was calculated using standard ballistics equations:
formula
where L is hop length, H is hop height, v is take-off velocity, g is acceleration due to gravity, and θ is angle of take-off. These two equations were solved simultaneously for θ to obtain the relationship:
formula
.

Five to ten hops were measured for each toad at each speed. Hop rates were measured by counting the number of hops taken during 30-to 60-s intervals of consecutive hopping at each speed. Two or three replicates were recorded for each toad at each speed and averaged to obtain the final value.

Statistics

Results in the text are given as means ± standard deviations. Standard correlation and regression analyses were used to examine the responses of and hop kinematics to changes in treadmill speed.

Gait transition

Toads used two gaits: walking and hopping. At slow speeds, walking toads had three feet on the substrate at all times, but toads often had only two feet on the substrate at fast walking speeds. Hopping, as defined by Emerson (1979), consisted of short jumps (<8–9 times body length). Toads showed a gait transition from walking at slow speeds to hopping at fast speeds (Fig. 1). At the slowest speed (0·09 km h−1), all the toads tested walked for some portion of the experimental period. Several animals walked nearly the entire time, but most walked and hopped intermittently. When speed was doubled (0·18 km h−1), the number of toads that walked halved. At faster speeds, the number of toads observed to walk continued to decrease until, at the fastest speeds (⩾0·36 km h−1), only a few toads walked during the experiments. At speeds ⩾0·36 km h−1 walking was confined to one or two steps taken to regain position after landing from a hop.

Fig. 1.

Gait transitions in response to treadmill speed in Bufo w. fowleri. The number of animals observed to walk during some interval of the experimental period, expressed as percentage of the total number of animals, is plotted as a function of treadmill speed.

Fig. 1.

Gait transitions in response to treadmill speed in Bufo w. fowleri. The number of animals observed to walk during some interval of the experimental period, expressed as percentage of the total number of animals, is plotted as a function of treadmill speed.

Oxygen consumption

Pre-exercise averaged 0·13±0·01 mlO2g−1h−1. Because toads explored the respirometer during the resting phase of experiments, this pre-exercise may be greater than the standard rate of O2 consumption.

During the first 5–10 min of exercise, increased to an approximate steadystate level , after which it usually remained constant but sometimes increased slowly, especially at higher speeds (Fig. 2). The values reported here are the average for the last 10 min of exercise.

Fig. 2.

Oxygen consumption (V·O2) of an individual Bufo w. fowleri (30 g) during 30 min of exercise on a treadmill at 0·09 (filled diamonds), 0·18 (open diamonds), 0·27 (filled triangles) and 0·45kmh−1 (open squares). Data for 0·27kmh−1 exemplify the response of an animal that refused to move during the first few minutes of exercise (see text).

Fig. 2.

Oxygen consumption (V·O2) of an individual Bufo w. fowleri (30 g) during 30 min of exercise on a treadmill at 0·09 (filled diamonds), 0·18 (open diamonds), 0·27 (filled triangles) and 0·45kmh−1 (open squares). Data for 0·27kmh−1 exemplify the response of an animal that refused to move during the first few minutes of exercise (see text).

Steady-state oxygen consumption increased significantly with speed between 0·09 and 0·27 km h−1 (Fig. 3A). The regression equation relating speed to for this range of speeds was (mlO2g−1h−1) = 3·87×speed (km h−1)+0·224 (r = 0·92). The slope of this line, often termed the minimum cost of transport (Cmin) (Taylor, Schmidt-Nielsen & Raab, 1970), had a 95 % confidence interval of 2·9–4·8mlO2g−1 km−1. The hypothesis of linearity across this speed range was not rejected (F[l,13] = 0·571, P>0·25). At speeds ⩾0·27kmh−1, was independent of speed (b =−0·016, P>0·25) and averaged 1·13 ± ·07mlO2g−1 h−1. The greatest average was attained at 0·27kmh−1 and was 1·17±0·1lmlO2g−1 h−1, nine times greater than preexercise .

Fig. 3.

(A) Steady-state V·O2(V·O2,ss) as a function of treadmill speed. Pre-exercise V·O2,V·O2,max and speed at which V·O2,max was attained (maximal aerobic speed, MAS) are shown. Filled circles indicate means. (B) Endurance (time to exhaustion) as a function of treadmill speed. Each open square represents a single trial. In B, points at 0·18 and 0·27 km h−1 obscure data for three animals at each speed; five animals were run at 0·27 km h−1. Trials were terminated after 120 min of exercise if exhaustion did not occur.

Fig. 3.

(A) Steady-state V·O2(V·O2,ss) as a function of treadmill speed. Pre-exercise V·O2,V·O2,max and speed at which V·O2,max was attained (maximal aerobic speed, MAS) are shown. Filled circles indicate means. (B) Endurance (time to exhaustion) as a function of treadmill speed. Each open square represents a single trial. In B, points at 0·18 and 0·27 km h−1 obscure data for three animals at each speed; five animals were run at 0·27 km h−1. Trials were terminated after 120 min of exercise if exhaustion did not occur.

Endurance

At low speeds (⩽0·27 kmh−1), six toads sustained locomotion for more than 2h (Fig. 3B). At 0·36 kmh−1, one of five toads was exhausted within 2 h, but even this toad sustained hopping for longer than 1·5 h. Endurance time decreased sharply at higher speeds to <10min at 0·63kmh−1.

Kinematics

The distance per hop (r = 0·42, 0·001 < P⩽ 0·005), height of hops (r = 0·60, P⩽0·0001), and angle of take-off (r = 0·67, 0·001 < P⩽0·005) increased with speed (Fig. 4A,B). Hop rate (hopss−1) increased with speed (r = 0·92, P⩽0·0001, Fig. 4C).

Fig. 4.

(A) Distance (open squares) and height (filled squares) of hops as a function of treadmill speed. The linear regression equation relating hop distance to speed is: distance = 7·1 × speed+10·3 (r = 0·42, 0·001 <P⩽0·005); the equation relating height to speed is: height = 3·9×speed+1·7 (r = 0·60, P⩽0·0001). (B) Angle of take-off during hopping as a function of treadmill speed. Take-off angles were calculated from distance and height data using standard ballistics equations (see text). The polynomial equation: angle = 68·6×speed+57·7×speed2+27·7 (r = 0·67, P⩽0·0001), was fitted to these data. (C) Hop rate as a function of treadmill speed. The equation relating hop rate to speed is: hop rate = l·5 ×speed+0·2 (r=0 ·92, P ⩽0 ·0001). Each point represents data for separate trials.

Fig. 4.

(A) Distance (open squares) and height (filled squares) of hops as a function of treadmill speed. The linear regression equation relating hop distance to speed is: distance = 7·1 × speed+10·3 (r = 0·42, 0·001 <P⩽0·005); the equation relating height to speed is: height = 3·9×speed+1·7 (r = 0·60, P⩽0·0001). (B) Angle of take-off during hopping as a function of treadmill speed. Take-off angles were calculated from distance and height data using standard ballistics equations (see text). The polynomial equation: angle = 68·6×speed+57·7×speed2+27·7 (r = 0·67, P⩽0·0001), was fitted to these data. (C) Hop rate as a function of treadmill speed. The equation relating hop rate to speed is: hop rate = l·5 ×speed+0·2 (r=0 ·92, P ⩽0 ·0001). Each point represents data for separate trials.

Aerobic response to exercise

Locomotion on the treadmill appeared to be natural. No obvious differences were seen in the manner of hopping when toads were unrestrained or on the treadmill. Our treadmill allowed the toads to make fine locomotor adjustments in response to speed (i.e. changes in gait and hop kinematics). Therefore, we are confident that our data reflect the aerobic metabolic support of normal, sustainable locomotion in B. w. fowleri.

The response of to speed in B. w. fowleri resembled what may be considered the ‘typical’ pattern of locomotor energetics for vertebrates: a steady, linear increase of with speed throughout the range of sustainable speeds, above which is constant and stamina decreases (Taylor et al. 1970; Taylor, 1977; Taylor, Heglund & Maloiy, 1982; Bennett, 1985). However, B. w. fowleri departed from the typical pattern in that although stamina decreased, it remained high at speeds above which was attained. For example, toads exercised at 0 ·45kmh−1 hopped for 50 ·90min (Fig. 3B). One possible explanation for this phenomenon is that hopping provides energy savings, presumably through the elastic storage of strain energy. Hence, as in kangaroos, hopping by toads at moderate speeds may be largely supported by aerobic metabolism supplemented by the recovery of strain energy. However, our data suggest that the contribution of strain energy is small (see below). Another possible explanation is that, although toads use anaerobic sources at these speeds, they might be able to accumulate a large lactate load without incurring fatigue. We made no measurements of anaerobic metabolism, but American toads, B. americanus, show elevated lactate levels at sustainable speeds (Taigen & Beuchat, 1984). Unfortunately, Taigen & Beuchat (1984) exercised their animals for only a few minutes. Sustained movement may be fuelled by aerobic metabolism after an initial contribution of anaerobiosis to support the first minutes of activity (Bennett, 1985). The mechanisms of stamina, fatigue and recovery are problematic in amphibians (Bennett, 1974; Cushman, Packard & Boardman, 1976; Fitts & Holloszy, 1976; Feder & Olsen, 1978; Hutchison & Turney, 1975; Hutchison, Miller & Gratz, 1981) and merit further investigation.

Minimum cost of transport

The cost of locomotion in hopping toads was no less expensive than running by other animals of a similar body size. The minimum cost of transport (Cmin) expresses the amount of energy required to move a given distance, and is frequently used for comparisons among taxa and modes of locomotion (Bennett, 1985). The minimum cost of hopping transport in toads is considerably greater than in running mammals (Taylor et al. 1982) or running lizards (John-Alder, Garland & Bennett, 1986). For example, Cmin of B. w. fowleri (=3·8mlO2 g−1 km−1) is 2·4 times greater than, and outside the 95 % confidence limits of, that predicted for a lizard of the same body mass (John-Alder et al. 1986). Moreover, the Cmin of B. w. fowleri is greater than those of salamanders (Ambystoma tigrinum, A. laterale, Desmognathus ocrophaeus, Plethodon jordani) run on a treadmill (Full, 1986; Full, Anderson, Finnerty & Feder, 1988). In fact, the Cmin of B. w. fowleri is 12·2 times greater than that of comparably sized A. tigrinum (Full et al. 1988).

Although sustained hopping by toads was at least as costly as walking or running by other animals, hopping at fast speeds may have been less costly for toads than walking at those speeds. This, of course, would account for the observed change in gait (Fig. 1). Unfortunately, because toads both walked and hopped at slow speeds and rarely walked at fast speeds, we could not discern differences in metabolic cost between the two gaits.

Efficiency of hopping

What fraction of the aerobic metabolic energy input is converted to mechanical energy output during locomotion in B. w. fowleri? The kinetic energy required to produce a hop can be calculated from the standard ballistics equations:
formula
where m is mass, and Ek is kinetic energy (Emerson, 1985). This estimate of Ek is minimal, because it ignores the energy required to move limbs (Cavagna et al. 1977). Power output can then be obtained as follows:
formula
In B. w. fowleri, mechanical power output increased linearly across the range of speeds tested (r = 0·95, P<0·0001; Fig. 5A). The energetic efficiency of locomotion was low in B. w. fowleri, ranging from 4 to 14 % (Fig. 5B) of the aerobic metabolic power input ( converted to units of Wkg−1). Efficiency increased at non-sustainable speeds (>0·27km h−1). However, at these speeds, anaerobic support for activity certainly became important. Therefore, aerobic power estimates for these speeds underestimate the total metabolic power input for hopping (aerobic+anaerobic) and overestimate efficiency. The low efficiencies at sustainable speeds are comparable to those for small mammalian hoppers (Biewener et al. 1981).
Fig. 5.

(A) Power output, in terms of kinetic energy, required to produce hops as a function of treadmill speed. Power output was calculated using standard ballistics equations and the equation for kinetic energy (see text). (B) Efficiency of converting aerobic metabolic energy input (V·O2,ss) to kinetic energy output (powerhop) as a function of treadmill speed. Powerhop data were the same as in A; V·O2,ss data were converted to units of W kg−1 using the relationship 1 ml O2 = 20·1 J. Stippling indicates the region of non-sustainable speeds. Each point represents data for separate trials.

Fig. 5.

(A) Power output, in terms of kinetic energy, required to produce hops as a function of treadmill speed. Power output was calculated using standard ballistics equations and the equation for kinetic energy (see text). (B) Efficiency of converting aerobic metabolic energy input (V·O2,ss) to kinetic energy output (powerhop) as a function of treadmill speed. Powerhop data were the same as in A; V·O2,ss data were converted to units of W kg−1 using the relationship 1 ml O2 = 20·1 J. Stippling indicates the region of non-sustainable speeds. Each point represents data for separate trials.

The low efficiency estimates for B. w. fowleri suggest that elastic recoil does not contribute a substantial amount of energy to sustained hopping. The efficiency of converting metabolic to mechanical energy ranges from 24 to 76% in kangaroos, depending on hopping speed (Cavagna et al. 1977). The maximal efficiency for the conversion of metabolic to mechanical energy by muscle contraction alone has been found to be 25 % (Hill, 1939; Gibbs & Gibson, 1972; Wendt & Gibbs, 1973). Thus, elastic recoil must provide the additional increments of energy used for hopping in kangaroos. In fact, as much as 70% of the kinetic energy that a kangaroo generates at the start of a hop can be recovered upon landing (Alexander & Vernon, 1975; Cavagna et al. 1977). Toads, however, showed low conversion efficiencies at all, and particularly sustainable, speeds (<14%, Fig. 5B), suggesting that most, if not all, of the energy required for sustained hopping was derived directly from muscle contraction.

These findings are consistent with our observations of sustained hopping by toads. Unlike kangaroos, they do not resemble a ‘bouncing ball or the action of a pogo stick’ (Dawson & Taylor, 1973) during sustained hopping. Instead, toads hop, return to a full ‘crouch’ (legs fully folded and abdomen touching substrate), and then hop again. Toads never hopped at a resonant frequency that remained constant regardless of speed, as expected in a system based on the alternate loading and unloading of a spring (McMahon, 1975). On the contrary, hop rate increased with speed (Fig. 4C).

Treadmill exercise compared with previous techniques

Maximal rate of oxygen consumption, is often used as an index of the upper physiological limit of sustained behaviour (Bennett, 1978, 1980; Bennett & Ruben, 1979; Taigen, 1983). From this perspective, covariation of with some aspect of behaviour has been taken as evidence for the adaptive significance of , such that high should potentiate higher levels of activity and, perhaps, a more complex behavioural repertoire than low (Bennett, 1978, 1980; Bennett & Ruben, 1979; Taigen, 1983). Consistent with this hypothesis, familial, species and even populational differences in among anurans have been correlated with differences in mode of locomotion (Bennett & Licht, 1973, 1974), foraging behaviour (Emerson, 1976; Taigen & Pough, 1983, 1985; Taigen, Emerson & Pough, 1982; Toft, 1980, 1981) and predator avoidance (Bennett & Licht, 1974; Hutchison & Miller, 1979).

Recently, these studies have been criticized because (1) the methods of stimulation do not elicit normal locomotion (Full, 1986), (2) different methods of stimulation may yield different values of for the same species (Hillman, Shoemaker & Putnam, 1979; Walsberg, 1986), and (3) some anurans attain higher levels during vocalization than during exercise in rotating chambers (Taigen & Wells, 1985; Taigen, Wells & Marsh, 1985). However, arguments advanced in support of previous measures of are that (1) the measures are reproducible for individuals, provided that the same method of stimulation is applied to all animals (Wells & Taigen, 1984; Walsberg, 1986; Walton, 1988), (2) increases with intensity of exercise, even though the exercise regime (i.e. rotation of respirometer jars) is not natural (Taigen & Beuchat, 1984; M. Walton, unpublished data), (3) during vigorous exercise covaries with the oxidative capacities of anuran locomotor musculature (Bennett, 1974; Putnam & Bennett, 1983; Taigen et al. 1985), and (4) during vigorous exercise, although not maximal, is a good predictor of (Taigen & Pough, 1985; but see Walsberg, 1986).

Our results validate measurements of maximum during non-normal locomotion in rotating respirometers. The reported here was similar to values obtained for toads that were overturned in closed respirometer jars at similar speeds (Table 1). However, if the cost of locomotion is at issue, our data suggest that more natural means of locomotion are necessary to yield acceptable results. The small cylindrical respirometers typically used in earlier studies do not allow the fine locomotor adjustments to speed found in our study. For example, in Taigen & Beuchat’s (1984) study, a ‘moderate’ level of activity encompassed speeds of rotation of 0·5–1·5 body lengths s−1. In our study, an equivalent speed range (0·09–0·27 km h−1) produced a change in gait, a 74 % increase in hop rate, a 35 % increase in hop height and a 98 % increase in (Figs 3,4).

Table 1.

Comparison ofV·O2,max in open-flow treadmill and closed-system respirometry

Comparison ofV·O2,max in open-flow treadmill and closed-system respirometry
Comparison ofV·O2,max in open-flow treadmill and closed-system respirometry

In conclusion, sustained hopping in B. w. fowleri was not less expensive than running in animals of a similar body mass, and was relatively inefficient at converting aerobic energy input into kinetic energy output. Thus, as in small mammals (Thompson et al. 1980), the energetic advantages associated with sustained hopping are probably small. Hopping at fast speeds may be less costly for toads than walking or running at those speeds. However, because toads both walked and hopped at all speeds and rarely walked at speeds above 0·18 kmh−1, the cost of walking and running transport was not quantified. The ability of toads to sustain hopping for long periods at speeds above the apparent MAS is problematic. Additional work concerning the anaerobic support for sustained hopping is required to resolve this point.

Others have argued that rapid escape from predators is the major selective advantage of saltatory locomotion in anurans and small mammals (Gans & Parson, 1966; Thompson et al. 1980). The anti-predator response of toads has been classified as ‘static’ (Taigen et al. 1982), but our experience has been that B. w. fowleri usually hops quickly away from pursuers. Bufo w. fowleri can hop at >1 km h−1, but only for about 30s (M. Walton, in preparation). Emerson (1978) suggested that anuran morphology allows the maintenance of constant acceleration over a range of body sizes. Hence quickness, not endurance, appears to be the primary benefit of saltatory locomotion in anurans.

We thank Martin Feder for assistance during the research and comments on the manuscript. We also thank Robert Full and Andrew Biewener for their helpful comments on the manuscript. Financial support was provided by US Public Health Service Predoctoral Training Grant in Genetics and Regulation 5-T32-GM-07197 to MW and by National Science Foundation Grant DCB84-16121 to M. E. Feder.

Alexander
,
R. Mcn.
&
Vernon
,
A.
(
1975
).
The mechanics of hopping by kangaroos (Macropodidae)
.
J. Zool., Land
.
177
,
265
303
.
Baudinette
,
R. V.
,
Nagle
,
K. A.
&
Scott
,
R. A. D.
(
1976
).
Locomotory energetics in a marsupial (Antechrnomys spenceri) and a rodent (Notomys alexis)
.
Experientia
32
,
583
.
Bennett
,
A. F.
(
1974
).
Enzymatic correlates of activity metabolism in anuran amphibians
.
Am.J. Physiol
.
226
,
1149
1151
.
Bennett
,
A. F.
(
1978
).
Activity metabolism of the lower vertebrates
.
A. Rev. Physiol
.
400
,
447
469
.
Bennett
,
A. F.
(
1980
).
The metabolic foundations of vertebrate behavior
.
Biosci
.
30
,
452
456
.
Bennett
,
A. F.
(
1985
).
Energetics and locomotion
.
In Functional Vertebrate Morphology
(ed.
M.
Hildebrand
,
D. M.
Bramble
,
K. F.
Liem
&
D. B.
Wake
), pp.
173
184
.
Cambridge, MA
:
Belknap Press
.
Bennett
,
A. F.
&
Licht
,
P.
(
1973
).
Relative contributions of anaerobic and aerobic energy production during activity in Amphibia
.
J. comp. Physiol
.
87
,
351
360
.
Bennett
,
A. F.
&
Licht
,
P.
(
1974
).
Anaerobic metabolism during activity in amphibians
.
Comp. Biochem. Physiol
.
48A
,
319
327
.
Bennett
,
A. F.
&
Ruben
,
J. A.
(
1979
).
Endothermy and activity in vertebrates
.
Science
206
,
649
654
.
Biewener
,
A.
,
Alexander
,
R. Mcn.
&
Heglund
,
N.
(
1981
).
Elastic energy storage in the hopping of kangaroo rats (Dipodomys spectabilis)
.
J. Zool., Lond
.
195
,
369
383
.
Cavagna
,
G. A.
,
Heglund
,
N. C.
&
Taylor
,
C. R.
(
1977
).
Mechanical work in terrestrial locomotion: two basic mechanisms for minimizing energy expenditure
.
Am. J. Physiol
.
233
,
R243
R261
.
Cushman
,
J. R.
,
Packard
,
G. C.
&
Boardman
,
T. J.
(
1976
).
Concentrations of lactic acid in neotenic and transformed tiger salamanders (Ambystoma tigrinum) before and after activity
.
J. comp. Physiol
.
112
,
272
281
.
Dawson
,
T. J.
(
1976
).
Energetic cost of locomotion in Australian hopping mice
.
Nature, Lond
.
259
,
305
306
.
Dawson
,
T. J.
&
Taylor
,
C. R.
(
1973
).
Energetic cost of locomotion in kangaroos
.
Nature, Lond
.
246
,
313
314
.
Emerson
,
S. B.
(
1976
).
Burrowing in frogs
.
J. Morph
.
149
,
437
458
.
Emerson
,
S. B.
(
1978
).
Allometry and jumping in frogs: helping the twain to meet
.
Evolution
32
,
551
564
.
Emerson
,
S. B.
(
1979
).
The ilio-sacral articulation in frogs: form and function
.
J. Linn. Soc. (Zool
.)
11
,
153
168
.
Emerson
,
S. B.
(
1985
).
Jumping and leaping
.
In Functional Vertebrate Morphology
(ed.
M.
Hildebrand
,
D. M.
Bramble
,
K. F.
Liem
&
D. B.
Wake
), pp.
58
72
.
Cambridge, MA
:
Belknap Press
.
Feder
,
M. E.
(
1986
).
Effect of thermal acclimation on locomotor energetics and locomotor performance in a lungless salamander, Desmognathus ochrophaeus
.
J. exp. Biol
.
121
,
271
283
.
Feder
,
M. E.
&
Olsen
,
L. E.
(
1978
).
Behavioral and physiological correlates of recovery from exhaustion in the lungless salamander Batrachoseps attenuatus (Amphibia: Plethodontidae)
.
J. comp. Physiol
.
128
,
101
107
.
Fitts
,
R. H.
&
Holloszy
,
R. H.
(
1976
).
Lactate and contractile force in frog muscle during development of fatigue and recovery
.
Am. J. Physiol
.
231
,
430
433
.
Full
,
R. J.
(
1986
).
Locomotion without lungs: energetics and performance of a lungless salamander
.
Am. J. Physiol
.
251
,
R775
R780
.
Full
,
R. J.
,
Anderson
,
B. D.
,
Finnerty
,
C. M.
&
Feder
,
M. E.
(
1988
).
Exercising with and without lungs. I. The effects of metabolic cost, maximal oxygen transport and body size on terrestrial locomotion in salamander species
.
J. exp. Biol, (in press)
.
Full
,
R. J.
&
Herreid
,
C. F.
(
1983
).
The aerobic response to exercise of the fastest land crab
.
Am. J. Physiol
.
244
,
R53O
R536
.
Full
,
R. J.
&
Herreid
,
C. F.
(
1984
).
Fiddler crab exercise: the energetic cost of running sideways
.
J. exp. Biol
.
109
,
141
161
.
Gans
,
C.
&
Parsons
,
T. S.
(
1966
).
On the origin of the jumping mechanism in frogs
.
Evolution
20
,
92
99
.
Gibbs
,
C. L.
&
Gibson
,
W. R.
(
1972
).
Energy production of rat soleus muscle
.
Am. J. Physiol
.
233
,
864
871
.
Herreid
,
C. F.
(
1981
).
Energetics of pedestrian arthropods
.
In Locomotion and Energetics in Arthropods
(ed.
C. F.
Herreid
&
C. R.
Fourtner
), pp.
491
526
.
New York
:
Plenum Press
.
Hill
,
A. V.
(
1939
).
The mechanical efficiency of frog’s muscle
.
Proc. R. Soc. Ser. B
127
,
434
451
.
Hillman
,
S. S.
,
Shoemaker
,
V. H.
,
Putnam
,
R.
&
Withers
,
P. C.
(
1979
).
Reassessment of aerobic metabolism in amphibians during activity
.
J. comp. Physiol
.
128
,
169
175
.
Hutchison
,
V. H.
&
Miller
,
K.
(
1979
).
Aerobic and anaerobic contributions to sustained activity in Xenopus laevis
.
Respir. Physiol
.
38
,
91
103
.
Hutchison
,
V. H.
,
Miller
,
K.
&
Gratz
,
R. K.
(
1981
).
The anaerobic contributions to sustained activity in the giant toad Bufo marinus
.
Comp. Biochem. Physiol
.
69A
,
693
696
.
Hutchison
,
V. H.
&
Turney
,
L. D.
(
1975
).
Glucose and lactate concentrations during activity in the leopard frog, Rana pipiens
.
J. comp. Physiol
.
99
,
287
295
.
John-Alder
,
H. B.
,
Garland
,
T.
&
Bennett
,
A. F.
(
1986
).
Locomotory capacities, oxygen consumption, and the cost of locomotion of the shingle-back lizard (Trachydosaurus rugosus)
.
Physiol. Zool
.
59
,
523
531
.
Mcmahon
,
T. A.
(
1975
).
Using body size to understand the structural design of animals: quadrapedal locomotion
.
J. appl. Physiol
.
39
,
619
627
.
Putnam
,
R. W.
&
Bennett
,
A. F.
(
1983
).
Histochemical, enzymatic, and contractile properties of skeletal muscles of three anuran amphibians
.
Am. J. Physiol
.
244
,
R558
R567
.
Taigen
,
T. L.
(
1983
).
Activity metabolism of anuran amphibians: implications for the origin of endothermy
.
Am. Nat
.
121
,
94
109
.
Taigen
,
T. L.
&
Beuchat
,
C. A.
(
1984
).
Anaerobic threshold of anuran amphibians
.
Physiol. Zool
.
57
,
641
647
.
Taigen
,
T. L.
,
Emerson
,
S. B.
&
Pough
,
F. H.
(
1982
).
Ecological correlates of anuran exercise physiology
.
Oecologia
52
,
49
56
.
Taigen
,
T. L.
&
Pough
,
F. H.
(
1983
).
Prey preference, foraging behavior, and metabolic characteristics of frogs
.
Am. Nat
.
122
,
509
520
.
Taigen
,
T. L.
&
Pough
,
F. H.
(
1985
).
Metabolic correlates of anuran behavior
.
Am. Zool
.
25
,
987
997
.
Taigen
,
T. L.
&
Wells
,
K. D.
(
1985
).
Energetics of vocalization by an anuran amphibian (Hyla versicolor)
.
J. comp. Physiol
.
155B
,
163
170
.
Taigen
,
T. L.
,
Wells
,
K. D.
&
Marsh
,
R. L.
(
1985
).
The enzymatic basis of high metabolic rates in calling frogs
.
Physiol. Zool
.
58
,
719
726
.
Taylor
,
C. R.
(
1977
).
The energetics of terrestrial locomotion and body size in vertebrates
.
In Scale Effects in Animal Locomotion
(ed.
T. J.
Pedley
), pp.
127
141
.
New York
:
Academic Press
.
Taylor
,
C. R.
,
Heglund
,
N. C.
&
Maloiy
,
G. M. O.
(
1982
).
Energetics and mechanics of terrestrial locomotion. I. Metabolic energy consumption as a function of speed and body size in birds and mammals
.
J. exp. Biol
.
97
,
1
21
.
Taylor
,
C. R.
,
Schmidt-Nielsen
,
K.
&
Raab
,
J. C.
(
1970
).
Scaling the energetic cost of running to body size in mammals
.
Am. J. Physiol
.
219
,
1104
1107
.
Thompson
,
S. D.
,
Macmillen
,
R. E.
,
Burke
,
E. M.
&
Taylor
,
C. R.
(
1980
).
The energetic cost of bipedal hopping in small mammals
.
Nature, Lond
.
287
,
223
224
.
Toft
,
C.
(
1980
).
Feeding ecology of thirteen syntopic species of anurans in a seasonal tropical environment
.
Oecologia
45
,
131
141
.
Toft
,
C.
(
1981
).
Feeding ecology of Panamanian litter anurans: patterns in diet and foraging mode
.
J. Herp
.
15
,
139
144
.
Walsberg
,
G. E.
(
1986
).
Comparison of two techniques for estimating the maximum aerobic capacity of amphibians
.
Herpetologica
42
,
389
394
.
Walton
,
M.
(
1988
).
Relationships among metabolic, locomotor, and field measures of organismal performance in the Fowler’s toad (Bufo woodhousei fowleri)
.
Physiol. Zool. (in press)
.
Wells
,
K. D.
&
Taigen
,
T. L.
(
1984
).
Reproductive behavior and aerobic capacities of male American toads (Bufo americanus): is behavior constrained by physiology
.
Herpetologica
40
,
292
298
.
Wendt
,
I. R.
&
Gibbs
,
C. L.
(
1973
).
Energy production of rat extensor digitorum longus muscle
.
Am. J. Physiol
.
224
,
1081
1086
.